Strava restricts API access, introduces monthly fee ahead of IPO
In brief
- Strava filed confidentially for IPO in January 2026, backed by Goldman Sachs and JPMorgan
- API access now requires monthly fees; fitness data banned for AI and machine learning
- Stricter guidelines limit developers to individual user data only, raising barriers for small builders
API Restrictions and Pricing
Strava updated its API agreement in November 2024 with stricter guidelines that already limited how developers could display data. The company restricted developers to displaying data for individual users only, cutting off broader integrations. Now, a monthly fee layer sits on top of those restrictions — a combination that creates friction for the ecosystem.
Strava explicitly banned the use of Strava data for AI or machine learning applications. That's a direct hit to health tech companies that rely on fitness data to build predictive models. The current API rate limits — 200 requests per 15 minutes and 2,000 requests per day — remain unchanged, but the new fee structure creates a floor cost that small developers may struggle to justify, especially if their tools serve niche audiences.
The IPO Context
Strava filed confidentially for its IPO in early January 2026, with Goldman Sachs and JPMorgan leading the process. The company's last known valuation was $2.2 billion, set during a funding round in May 2025. CEO Michael Martin has signaled that the IPO is meant to fuel growth and acquisitions, with a particular focus on capturing Gen Z users.
Behind the scenes, the numbers tell a different story. Strava's annual recurring revenue is estimated to be approaching $500 million, with premium subscriptions accounting for roughly 80-90% of total income. The company is already highly profitable on a per-user basis. The API fees aren't about unlocking new revenue streams — they're about control.
What This Means for Developers
The restrictions hit hardest on smaller builders. Fitness data is enormously valuable for health tech companies building predictive models, but Strava has now made it harder and more expensive to access. Individual user data only, no AI training, and now a subscription cost. It's a three-layer wall.
This strategy may backfire. Developers who've built on Strava's API have been a source of organic marketing and ecosystem value. Pricing them out risks losing that advantage. The question isn't whether Strava can enforce these rules — it can. The question is whether it should, and what it costs in terms of platform innovation and loyalty.
